Incorrect.
It was a reduction in what the expected GDP would have been with a Remain vote.
It was not a reduction in GDP overall.
I took that into account. The prediction of GDP without Brexit wasn't particularly rosy and the pessimistic forecast of the change in GDP was so hugely negative that it sent the overall GDP forecast negative too.
Anyway, we should consider actual forecast values, these are bad enough. If you wish to compound that by forecasting a change to an existing forecast you're just making things even more difficult to understand and making it impossible to check the forecast accuracy. When reality turns out different to the compounded forecast, was it the first forecast that was wrong, or the second one, or maybe both of them?